|
Dirty-South Blues Harp forum: wail on! >
THE COMB DEBATE REVISITED: SPAH Test
THE COMB DEBATE REVISITED: SPAH Test
Page:
1
Brendan Power
47 posts
Jul 13, 2010
5:33 AM
|
As most of you will know, Vern Smith has conducted a couple of public tests over whether listeners in an audience could detect differences in sound between harmonicas with different comb materials. The result was conclusive: they can't. Vern has offered a wager of $1000 to anyone who can prove that result wrong, and no-one has taken up the offer. That speaks for itself!
I fully accept the results of Vern's audience tests, but now want to investigate a different issue: what differences (if any), can players/very close listeners detect? This issue still arouses passionate feelings, as is shown by the way many players are very choosy about their favourite comb materials. Should they bother? Hopefully this test will shine some light on this fascinating and still murky area.
There are different ways to test this issue. I'm adopting a passive approach, simply collecting data, not seeking to prove anything one way or the other. I'm very happy to say that Vern Smith has kindly agreed to lend his valuable experience and expertise to my test, and in the future may follow up with a differently-focussed test of his own. Here is a description of the 2010 SPAH Test:
-----------------
THE COMB DEBATE REVISITED: Player, Close Listener & Spectrograph Test
Co-Conducted by Brendan Power & Vern Smith. Custom combs supplied by Chris Reynolds
The aim of this test is to see what can be discovered about the sounds of a harmonica with different comb materials, when listened to up-close under controlled conditions by three parties: 1. Players 2. Close Human Listeners 3. Microphone and Spectrograph. One vexed issue we hope to shed light on is "Can the player perceive differences in comb materials?" In doing the testing, we will record spectrograph images of many common comb materials, and there could be some interesting ancillary data (eg. possible tonal differences between embouchures: Tongue Blocking vs. Lip Pursing vs. U-Blocking).
We will use only one set of diatonic reedplates and coverplates on all combs, from a Hohner Marine Band Deluxe. Vern has made a special bracket which will enable fast comb changes and mask the varying weights of the different combs. The comb smell/feels will be masked so that the different materials cannot be sensed by lips/tongue.
We will be testing eight materials which are commonly used in diatonic harmonica combs: Pear Wood, ABS Plastic, Diamondwood, Corian, Acrylic, Brass, Bamboo, Aluminium Alloy. Aside from the stock Hohner combs, the identically sized/shaped custom combs will be made to exacting specifications by Chris Reynolds. They will be available for sale after the test.
There will be six test players. They will be blindfolded and wear ear mufflers between each test (so they cannot listen to the other testers before/after their test). They will be respected names who have different styles and embouchures (tongue block/lip purse/U block). For the spectrograph readings, two single test tones of 3 seconds each will be taken for each comb: a blow and a draw on hole 4. For the test playing, each player will choose a favourite selection of phrases which they duplicate to the best of their ability on every comb. Each playing test will last 20 seconds. Players will note their impressions immediately after each test, in isolation. The players will not confer or communicate with each other in any way during the testing procedure.
The three Close Human Listeners (within 30 cm away from the played harp) will also be blindfolded for each test, and not permitted to confer/communicate with each other. They will also note their impressions immediately after each test, in isolation.
The data from the Players and Close Human Listeners will be collected via simple unbiased multi-choice questionnaires. It can later be compared to the Microphone/Spectrograph data for each comb. The data will be made freely available to all after the test.
An audience is invited to attend the test but must remain silent when each test is conducted, and must not influence the testing in any way.
-----------------------------------
No doubt there will be many suggestions/questions about the procedures for this test! Rest assured that Vern, Chris and I have already engaged in LONG email discussions over several months examining all the quirks/ramifications/implications of different elements of the testing procedure, and have come up with what we collectively believe to be the best possible process for a two hour test at SPAH. We are confident that our procedure and protocols will be sufficiently rigorous, comprehensive and balanced to produce valuable data that will add considerably to the long-standing debate over whether comb materials affect harmonica tone. The aim of this particular player-based test is not to prove anyone right or wrong, but to increase the body of knowledge about this fascinating, hotly-contested issue. At present, it is purely a matter of (widely differing) subjective opinion. Hopefully the results of this test will start to give the player-opinion debate a more factual basis, and future tests can build upon it to reach a widely accepted consensus.
Brendan Power
WEBSITE: http://www.brendan-power.com
YOUTUBE: http://www.youtube.com/BrendanPowerMusic
Last Edited by on Jul 13, 2010 5:44 AM
|
MrVerylongusername
1127 posts
Jul 13, 2010
6:08 AM
|
I'm very glad you are doing this. The procedure sounds far more scientifically robust than anything else I've heard about.
|
HarpNinja
545 posts
Jul 13, 2010
6:20 AM
|
It doesn't surprise me that most in the audience wouldn't have a clue as there are so many factors that can skew what they hear.
However, I am now convinced that having a good comb makes the harp easier to play. This may be more about the seal of the comb to the reedplates than anything else. I know that is slightly off topic, but custom combs or modded combs or really good factory combs make the harp more air-tight, play with more presence and sometimes change the player's embouchre.
Just yesterday I got a Low E Marine Band. I have some Buddha combs in the mail, so for now I just sealed the wood comb and reassembled using the nails on the reedplates and screws for the cover plates.
I was very lazy and only flat sanded the comb (then rubbed a little chapstick on it). I didn't mess with anything on the reedplates and in fact probably warped them a bit in getting them off. I am also not so sure the nails are as snug. Regardless, the harp plays better now and has a bit more bite. The middle of the harp is especially easier to play.
This is as close to a "stock" sealed MB as I could get. There is no reedwork whatsoever.
But I'd still predict that tonally, the comb doesn't mean as much as say the reedwork or cover plates.
----------
Mike
|
5F6H
244 posts
Jul 13, 2010
7:21 AM
|
Considering the myriad differences in all other aspects, apart from comb material (reed material & dimensions, coverplates, comfort, tuning), with regards to the regular "off the shelf harp"...not being funny, but what exactly is the point of the "test"?
If it's "proven" (within the confines of the test perameters and the limits of the methodology) for instance, that a treated pearwood/composite/compressed banana skin comb sounds the same as untreated pearwood, it doesn't mean that people who have swelling issues will stick to untreated pearwood.
There are plenty of perameters that dictate what kind of harp you buy (& subsequently what comb material) beyond specifically what the comb is made of.
Playing devil's advocate here, but is the fact that Vern's $1000 has never been claimed due to the fact that his data is robust, or that no one is really so bothered about comb material that they are prepared to risk $1000 & end up in a long, drawn out squabble about methodology & whether all parties can agree on who is right/wrong?
A live test featuring one set of reedplates brings it's own problems, even folks with reasonably good hearing have trouble retaining detailed nuances of tone & timbre for even very short periods of time. As soon as they hear a new sample it typically overwrites the old sample unless there is something very specific that one, or the other, lacks or has too much of. Listening, in this respect, is an aquired skill people (including many talented musicians) are not born with it. Multitracking a good recording of a set piece & switching between tracks would be a more realistic "test".
It strikes me that we already know that players sound different, so why use more than one player?
It's analogous (to my mind & my sphere of interest) to "does a brand of tube make a difference"? Let's say in my amp, with my mic, I can tell one brand of preamp tube from another & given the 2 options, that I could detect which was which, with a 90% success rate. If someone else fails to do so, it doesn't mean that there is no difference, just that they might not be so good at listening.
Now change the player and it gets more tricky, I might be able to tell a difference reliably, but not necessarily which was which...the more players & brands the harder it gets so in the end, over a dozen tube brands, accross 6 players I wouldn't stand a good chance of picking out which was which...but it wouldn't prove that there is "no difference", you might just say that as long as everything is working within accepted perameters, that there "isn't usually enough difference to worry about/prevent a good performance"... a different standpoint entirely.
So at the end of the day, when a customer tells me that he likes brand A over brand B, contrary to my suggestion perhaps, I give him brand A at the market rate & send him away happy...rather than waste an entire man day (which I wouldn't get paid for) to "prove" otherwise.
I don't doubt the skills of the craftsmen involved, nor their integrity, but I struggle to see how a meaningful test (despite the irrelevance of the result) can be executed.
Last Edited by on Jul 13, 2010 7:40 AM
|
Brendan Power
48 posts
Jul 13, 2010
5:36 PM
|
5F6H: You misunderstand the aim and methodology when you say "It strikes me that we already know that players sound different, so why use more than one player?" and "...across 6 players I wouldn't stand a good chance of picking out which was which."
It's the Players' opinions we're examining, so of course you need more than one! As for the Close Listeners, they're not trying to pick out comb types from a list, as you seem to infer...
As the combs are switched they, (blindfolded, like the players), will simply note down if the harp sounds bright/dull, loud/soft etc. Each descriptive adjective will have three levels: 1. 'Not Very'; 2. Average' and 3. 'Very'. We can use those numbers to acquire overall scores. So at the end, we will get a graph of opinions of each comb type from 6 players and three close listeners. That should yield some useful data, especially when compared to the spectrograph images.
Yes, all the tests will be recorded and put up online so anyone can make up their own minds from what they hear, in addition to the impressions of the players ands close listeners on the day.
The point? Just increasing our knowledge of something that's dear to a lot of players' hearts, that's all. Isn't that worthwhile in itself?
|
MP
689 posts
Jul 14, 2010
1:27 AM
|
you know,
i can't tell the difference between different makes and models of harps if other people are playing them.
well, maybe i could if it was an extreme bender, but it's doubtful. funny name for a harp.
i can't tell if they are ET or compromise, or 7 or 19 limit JI.
when i customize my harps, the crossover sounds like the deluxe, that sounds like the MB, that sounds like the SP20.
my tounge and hands tell me what im playing.
i don't get it brendan, sorry.
----------
MP
hibachi cook for the yakuza
doctor of semiotics
superhero emeritus
|
Brendan Power
49 posts
Jul 14, 2010
3:56 AM
|
MP: Many players care a LOT about what comb they use - not just the shape and feel in the mouth, but what it is made from too. They will often pay a premium to have a particular material for their custom comb - whether it's sealed pear wood, titanium, corian, dymondwood, acrylic, brass or whatever.
Players (and manufacturers) frequently make claims about the sounds from different comb materials. Common ones are "Wood is warm", "Steel is bright" etc etc.
At present, these are simply subjective, unverified claims. If the effects ARE there, they should be measurable and quantifiable in a scientific sense: that would be very interesting to know, for players and comb makers alike. Or perhaps they are not really there at all, just in peoples' minds - that would be interesting too.
My SPAH test will hopefully provide some useful data that can help towards answering these questions.
|
MP
693 posts
Jul 14, 2010
4:36 AM
|
yes brenden, THEY must care a lot!
i went to buy some maple combs for some SP20s one week after i saw them and they went from $15 to $20 and by the time i decided, they were sold out.
ditto with most of the other combs for GMs etc.
i wouldn't mind some corian and dymondwood, but for me it's aestetics not comfort.
once the harp is in my mouth in a matter of seconds i fail to notice sharp edges. depth etc. unless it's a chromatic.
perhaps comb material does affect a harmonicas tone.
i think the listening device to measure such differences would have to be very sophisticated; like a doppler.
our ears are sophisticated devices-well, some peoples are, but i don't think they can be trusted for such a scientific undertaking.
----------
MP
hibachi cook for the yakuza
doctor of semiotics
superhero emeritus
|
chromaticblues
152 posts
Jul 14, 2010
5:13 AM
|
I have always wanted to go to SPAH. That test sounds interesting. I don't have to agree with the findings, but I'm interested. I would like to do the blind fold test myself and have one of my own custom marine band in the line up. I don't mean that in an arguementive way! I would just like to do that test myself and not involve anyone else(except someone to hand me six different harps). Of course I'm not betting $1000. I think I can tell the difference. I think using very skilled players that play Marine Bands might change the results if the test was conducted the way I described! But as someone already stated; $1000!! It don't make me no never mind!
|
N.O.D.
30 posts
Jul 14, 2010
7:20 AM
|
Hi Brendon:) you mention
Players (and manufacturers) frequently make claims about the sounds from different comb materials.
Common ones are "Wood is warm", "Steel is bright"
I Find this claim to be true from my own experience
I used to play Lee Oskars i bought a set of Susuki Pro masters for my farrari set of Harps,
after playing LO's for years the first thing i noticed with the Pro Masters was that they where very much brighter than the LO's and Projected sound much more
like going from acoustic guitar to Dobro guitar the sound was louder,
i then bought a set of Pure Harps i thought hay i got some aluminium combs why not some wood combs
i believe these have the same reed plates as the Pro Masters
there is a remarkable difference in tone as you say warmer or not as Bright as the Pro Masters
At present, these are simply subjective, unverified claims.
If the effects ARE there, they should be measurable and quantifiable in a scientific sense:
you have a close relationship with Susuki Harmonicas
im sure they must have done some sort of controled tests, or you could say scientific tests for the introduction of the Manji Harp,comaring them to the comb materials,
I liked the concept of the Manji with the compasite comb not because i thought it may make a difference in tone but of it solid construction,
I'm hard core SP-20 the only thing i don't like about plastic combs is ther construction what i call a skelatel construction ie semi hollow ribs, ridges and hollowed out,
i would be Happy with a Solid Comb SP-20 im not worryed about resonance but i would suspect there would be some or maybe say sustain im not sure we will leave that up to your tests,
as for Blowing the Myth out of the water does it realy matter what material the comb is made out of,
well lets hope for the sake of the Persons earning a quid from custom combs, that there is a difference
because if there isn't why bother with custom Combs,
About those Close listeners I'm not being rude or offensive but it would be great if you where able to offer a position to a Blind Muso us sight impaired people we got some mighty fine ears on our heads
i used to be 95% blind in the left eye and not to good on my right side relyed heavily on my ears i have to tell my Bro his guitar is out of tune i will have to stop say mate your guitar is out, and it is mostly only one on 1 or 2 strings slightly out he say it was just out,
Thats just a thought it would be a good Public relations thing to i guess,
about that thousand buck thing the stakes are to high
sort of like saying i will prove you wrong but not willing to be proved wrong,
by raising the stakes very High $1000 persons are unwilling to match this bid the guy knows it so he gets no takers,
this in his own mind confirms his belief that he is right lower the stakes you will get takers:)
----------
Thats my Lot might sound like rot
but thats what i got
Good Onya Mate
|
MrVerylongusername
1131 posts
Jul 14, 2010
7:44 AM
|
To those people asking why? - there doesn't have to a "why"; science can sometimes just be about pushing boundaries of knowledge. Nethertheless in this case Brendan has given a very convincing reason; challenging subjectivity. With the current state of knowledge it is still possible for people to claim that wood=warmer etc... There is precious little (if no) empirical evidence to challenge or support those claims.
My suspicion is that players are hearing what they want to hear, but I would be very happily corrected if the science proved otherwise. People have biases, but (good) science is neutral.
Wood=organic=traditional; does that necessarily also =good? plastic=manmade=modern; so often in this world those imply inferior . Does that logic apply here? Can we 100%, with our hands on our hearts claim our perception is not even slightly clouded by these expectations?
As for comb choice. I am the opposite of MP. To me it is not about perceived tone, but whether it feels like I'm playing something comfortable or whether my tongue feels like it's sliding across a cheese grater.
I U-block, so my tongue is always in contact with the comb, I've had very negative experiences with Marine Band comb swelling. Plastic is the winner for me.
There will always be preferences. This research isn't going to sound the death knell for wood or plastic combs.
Mark suggests that maybe people don't really care enough to prove Vern wrong. Perhaps that's true.
$1000 dollars is high enough to discourage an individual whose methodology was not 100% watertight, but it's small change/low risk to the big manufacters. Why haven't they challenged Vern? - probably because they think the status quo benefits them.
I actually can see some practical reasons why resolving this issue might be good for players and manufacturers.
For instance, I would love to see more manufacturers doing what Seydel are doing with the 1847 - offering the same model with different comb options; wood, sandwich style plastic, recessed reedplate style plastic etc... without having to go to a customiser or do a DIY job to get it. I'd love Golden Melodies to be offered with a sealed wood comb; I'd be interested to try a Marine Band Crossover with a plastic comb. Personal choice. Surely more options are better? But I see a barrier in the way that Hohner chooses to market their harps - they are really fixated on the historic/tradition thing (not surprisingly - I guess it sells harps by the truckload).
If there was proof that comb material effected tone, their wood=traditional=warm stance would be vindicated and they could offer a choice and push the tonal benefits: "Golden Melody - now with the warm, traditional sound of a wood comb!"
If the proof went the other way, they could still be onto a winner: "Same Marine Band sound, but now with the comfort and durability of an ABS comb!"
It's probably a pipe-dream. I know there's all the issues of production/tooling etc... but like I said - Seydel are already doing it and it works for them.
I do wonder sometimes if the developers at Hohner have already conducted experiments like these i.e. proper blinded tests that do away with any expectation on the part of the observer. If they have they are keeping it to themselves. Maybe they are worried about changing their very conservative approach to marketing? When Hohner asked Howard Levy and Joe Filisko to comment on the Crossover, did they know what they were playing? No disrespect intended to either, but even great people have expectations and pre-conceptions.
Last Edited by on Jul 14, 2010 7:48 AM
|
Brendan Power
50 posts
Jul 14, 2010
7:53 AM
|
You make many good points, MVLU.
Just to clarify, my 2010 SPAH test has nothing to do with Vern's $1000 challenge. That is about what an audience can detect, this test is about what players/close listeners detect.
Will you be at SPAH? I need a good U-Blocker to be one of the testers (so as to cover all three embouchures). If you won't, or don't think you're the man for the job, can you suggest a good U Block player or two I can approach? Thanks, Brendan
|
MrVerylongusername
1132 posts
Jul 14, 2010
8:11 AM
|
That's a great offer! I wish I was going Brendan, but I'm in the UK and have no money to fly Stateside right now. One day maybe, when the kids have grown up and got their own jobs...
I understand that what you're doing is different to Vern's $1000 challenge, but it is hard to consider the audience perception (Vern's challenge) without considering the player's point of view. What we have is some strong evidence for one side of the coin and this experiment to (hopefully) firm up the evidence for the other side. I was just curious about why the big companies are leaving it to the amateurs (from an R&D perspective: obviously not from a harp playing one!) and individuals to resolve the whole issue - like I say it probably suits them not to have an answer, or not to publicise it if they do.
Sorry I don't know any other U-blockers, but don't forget to ask amongst the chrom players too.
Last Edited by on Jul 14, 2010 8:16 AM
|
Jim Harris
31 posts
Jul 14, 2010
10:37 AM
|
Brendan -- My suspicion is that the cover design plays a far greater role in the tonal quality than the comb. I wish you were able to have a standard and non-standard (more open back, and perhaps thicker) cover to check the sound difference. Of course that would lengthen and complicate your test, and detract from the comb analysis. I notice the Suzuki Bluesmaster and Promaster seem louder and brighter than SP20 and MB, and I can't help but notice the cover shape and more open design.
NOD -- Even if the comb play little or no role in affecting the tone, I think there is still valid reason to use a custom comb. Not the least is the visual -- seeing at a glance which harp you want. And often it is comfort factor with less edge. I'm curious about those raised/rounded tines for the mouth fit. I doubt we need to worry about the custom comb makers losing much business.
Last Edited by on Jul 14, 2010 10:39 AM
|
MP
697 posts
Jul 14, 2010
1:38 PM
|
i still want corian, dymond wood, and maple ala some seydels.
i'll U block like i was born to do it, and shamelessly repeat whatever unsubstantiated claims necessary if they award me with these fine combs.
----------
MP
hibachi cook for the yakuza
doctor of semiotics
superhero emeritus
Last Edited by on Jul 14, 2010 1:41 PM
|
jim
224 posts
Jul 15, 2010
3:46 AM
|
hey!
I advise you to make 3 special combs for some harp model - one wooden, one from metal, one from plastic. All "sandwich" style (i.e. plastic like 1847silver).
And here's the trick:
the front of all coms should be the same (say, with a metal outline) - I suggest that to make the test 100% accurate. The player will not FEEL any difference that way when playing with eyes closed - they would all feel like a metal comb in the mouth, despite the other 99% being metal/wood/plastic.
Would be fun to check the opinions of the players that way.
|
5F6H
246 posts
Jul 15, 2010
5:32 AM
|
Jim. Hohner already did that with the Marine Band style orchestral 10 hole harp, it had a metal front to the pearwood combs...what if the metal inlay itself, rather than the comb material itself, insulated by the metal, affected tone?
Brendan - Sure, I hear you on the objective of your test. Don't get the idea that I'm not intrigued...playing devil's advocate to some degree, but also think that simplifying the test, based around the commonly available combs might be more productive...a sample of 6 candidates is interesting, but I'd expect a bigger sample, in terms of players & maybe less, or rotated, samples for the comb materials (especially the more esoteric materials) for a more robust comparison test...but then, that would cost money & would really be in the realms of what a manufacturer should be coughing up for. In short, if this were market research, I don't think anyone would launch/revise a product based on such a small sample.
Since there will be plenty of players on hand it seems a shame/wasted opportunity to only ask 6 for their opinions...of course to do more players means more materials & more time &/or a couple of guys stripping cleaning & rebuilding harps between tests...
Spectrograph is interesting, but is really just icing on the cake.
I remember a friend of mine giving me a plexiglass comb to try in comparison with wax sealed pearwood. I had to go through a few harps of the same key (& build materials) until I found 2 that sounded as similar as I could, then changed the combs - the 2 most similar harps were still the most similar despite the comb material change, there was a bigger difference between the remaining harps of seemingly "identical" construction. Not suggesting that this is any kind of a precedent, it's just an anecdote.
Last Edited by on Jul 15, 2010 9:27 AM
|
Honkin On Bobo
345 posts
Jul 15, 2010
7:30 AM
|
I subscribe to the BarbequeBob theory of harp playing which states that over 90% of the sound out of the harp is due to the player not the gear. To that end, I don't really think about comb material much. But it is nice to know that someone is looking into this in an objective manner.
With a light wallet, here's to hoping the results show it doesn't matter what the comb is made of. If I find out I need the special Tiki wood blessed by the ancient tongan shaman to sound good, I'm switching to kazoo.
Good luck with the study Brendan.
Last Edited by on Jul 15, 2010 7:44 AM
|
barbequebob
1024 posts
Jul 15, 2010
11:07 AM
|
As far as the differences, it's more likely to be subtle rather than dramatic from where the players' ears are concerned. From the audience end, regardless of comb material or covers, it's not gonna make much of a real difference.
On the other hand, everyone's ears are different, as some are more sensitive to highs, lows, etc., and whichever debate it is or how it ends up (and that's gonna be argued forever), that's gonna be tough to really be truly conclusive realistically.
----------
Sincerely,
Barbeque Bob Maglinte
Boston, MA
http://www.barbequebob.com
CD available at http://www.cdbaby.com/cd/bbmaglinte
|
MP
705 posts
Jul 15, 2010
11:49 AM
|
my point exactly bob.
@5F6H, not rubbish at all. perhaps i should have made my point clearer. the differences would be so subtle; and our ears are not equiped with even the most basic recording devices, (like a pitch wheel) and strobe meters to record actual tonal readings, as to make any outcome subjective to the point of uselessness.
on the other hand, i trust my ears on playback devices. in blindfold tests iv,e done, analog beats out digital every single time without fail.
----------
MP
hibachi cook for the yakuza
doctor of semiotics
superhero emeritus
|
gene
524 posts
Jul 15, 2010
3:16 PM
|
I'm inclined to say that humans ears are not needed for the test. The spectrogragh should tell all. I'm also wondering if a machine should do all the "playing." Assuming the combs are all identicle in shape & size, all harps would be played identically. The machine would alas eliminate any possibility of the test being flawed due to comb masking because the combs would not have to be masked. I think the machine should blow and draw all ten holes, too. Higher or lower pitches might affect the combs to different degrees.
As I'm writing this, I am realizing that if a spectrograph shows a difference, the question would remain whether a human could detect the difference shown, but I get the feeling it would already be documented in medical science just how much difference a human ear could detect.
Last Edited by on Jul 15, 2010 3:33 PM
|
MP
711 posts
Jul 15, 2010
3:20 PM
|
i can tell the differences in 5 brands of scotch, or between 7 and 12 year old.
----------
MP
hibachi cook for the yakuza
doctor of semiotics
superhero emeritus
|
Honkin On Bobo
346 posts
Jul 15, 2010
4:12 PM
|
Uh....MP.....I hereby volunteer myself for any experiments you wish to conduct.
|
MP
714 posts
Jul 15, 2010
5:16 PM
|
Bobo,
you'll have to sigh a waver. i'm sure you understand. i'll add that you are a brave and selfless man. YAY for science!!!
----------
MP
hibachi cook for the yakuza
doctor of semiotics
superhero emeritus
|
MP
716 posts
Jul 15, 2010
9:36 PM
|
@Bobo and groyster,
then it's settled gentlemen!
we shall loan our bodies to holy science and conduct our experiments in Scotland!
----------
MP
hibachi cook for the yakuza
doctor of semiotics
superhero emeritus
|
5F6H
249 posts
Jul 16, 2010
1:45 AM
|
Gene & MP - What Brendan is realy suggesting is called "sensory testing", he is looking for percieved differences & opinions on such. This kind of testing is carried out every day of the week on consumer products. The "science" applied is based more on ensuring that the tests are reproducible & unbiased. There are different methodologies depending on the level of detail required.
Brendan appears to be looking more for how people feel about what differences that may (or may not) be detected. Not so much a clinical, so called "scientific" (a word that is bandied about freely without much real understanding of what scientific testing is - e.g. "if we use a machine, then it's scientific test surely?") to determine whether there is, or isn't a difference (what Vern was attempting).
Brendan, I really thing the one set of reed plates is likely to foul things up, surely there is time to fine tune & cherry pick enough sets to avoid the down time between materials? Additionally, reed plate sets could be rotated between the comb materials, between different candidates, to eliminate any bias.
Chefs, whisky blenders, somelliers, artists, amp builders, recording engineers, musicians etc, etc, all use their senses as the final verdict.
|
MrVerylongusername
1133 posts
Jul 16, 2010
2:48 AM
|
Different reedplates introduces an extra variable. To eliminate that variable each player will have to play each set of plates with each comb. I don't see the point. It just complicates things unnecessarily.
|
5F6H
250 posts
Jul 16, 2010
4:54 AM
|
VLUN - Reedplates & cover plates can be cherry picked & selected by ear prior to the test. Even if they are not exactly "identical" (but hopefully, close enough to be reasonably similar) then each player does not have to try each comb with each set of plates, but player #1 has comb #1 & reedplate #6
player #2 has comb #1 & reedplate #3
player #3 has comb #1 & reedplate #1...
...& so on, so that reedplates are rotated accross all the candididates, so accross the test there shouldn't be any bias attributable to the plates themselves, with respect to the combs...but if the results follow the plates, then this is traceable too.
I have just read on harp-l that Vern is saying that all combs will have a barrier layer (so all feel the same) so as not to give the game away...will this barrier layer itself affect the acoustic properties of the comb? If not, how not?
Last Edited by on Jul 16, 2010 4:55 AM
|
MrVerylongusername
1134 posts
Jul 16, 2010
5:16 AM
|
The masking layer might affect the tone, but it's effect will be introduced into each trial, thus cancelling itself out.
Your scenario assumes that each player's hearing and perception (and those of the person "cherry picking" the reedplate) are equal. Reedplate #1 may have some tonal character that only player #1 can detect (perhaps because the others have stood next to a gorilla bashing his crash cymbal for too many years.) If the other two players cannot detect it, there is a risk that the effect is attributed to the comb and not the reedplate - the only way to prevent that would be to give player #1 the opportunity to play the same comb with a different reedplate.
Players' hearing acuity is one variable, thus each comb should be played by each player.
Comb material is another variable, thus each player needs to play each comb.
Actual reedplate would be a third variable requiring each player to play each reedplate / comb combination.
I do agree that the sample size is a little small though.
Last Edited by on Jul 16, 2010 5:34 AM
|
5F6H
251 posts
Jul 16, 2010
5:53 AM
|
VLUN "The masking layer might affect the tone, but it's effect will be introduced into each trial, thus cancelling itself out." Indeed, but if the masking layer is not present in any available combs AND it has an effect acoustically, then the test has limited real world value. It will only be a test of combs with the masking layer.
Yes, agreed, players hearing acuity is vary variable (some of the best players I know have excellent relative pitch & great musicianship, but frankly less than perfect hearing after playing with loud bands for years), that's why the sample, in terms of players needs to be bigger. Whether the 6 players in question are good players or not, is not as relevant as their hearing acuity...unless this is also being tested prior to the comb test? Hence, the more players on hand, the more data & the less impact a particular players acuity, or lack of it, will have.
Over a sample where every reedplate is tested with every comb an equal number of times, the effect of the reedplate will have little bias on the comb results (even if it does, this will be traceable if correctly executed). And as I said, there should be time to pre-select the reedplates (all of an identical design) to eliminate variances as much as possible in the first place.
At the moment, the sample in terms of players is so small as to just be a test of what 6 players think, there will be lots of players available, some who may have excellent hearing, the more players, the more robust the data.
Easy to say, I know, because I'm not going to be there, organising, stripping, cleaning, rebuilding, rotating harps...but if it's worth doing & publishing, then it's worth doing something that will stand up to rigorous scrutiny...'cos that's what it is going to get.
|
MP
720 posts
Jul 16, 2010
12:02 PM
|
5F6H,
since you got all argy bargy and said "RUBBISH",
you're not allowed to join in the single malt tests w/ groyster and Bobo in scotland. so there.
----------
MP
hibachi cook for the yakuza
doctor of semiotics
superhero emeritus
|
Brendan Power
52 posts
Jul 17, 2010
8:51 AM
|
MVLUN wrote "Different reedplates introduces an extra variable. To eliminate that variable each player will have to play each set of plates with each comb. I don't see the point. It just complicates things unnecessarily."
That's right. It would invalidate the test, or make it incredibly long, to have different reedplates.
5F6H wrote "but if it's worth doing & publishing, then it's worth doing something that will stand up to rigorous scrutiny...'cos that's what it is going to get."
Certainly. I believe the testing process will stand up to scrutiny and deliver worthwhile human impressions and spectrograph data. However, that's for you and everyone else to judge afterwards. All the data including, sound files, spectrograph images and questionnaires will be put online so anyone can check it and make up their own minds.
I'm not trying to prove anything one way or the other, just gather impressions and data in as careful and unbiased way as I can, and offer it to others to judge. Hopefully others will do further tests in the future, to add to the available information.
|
5F6H
252 posts
Jul 19, 2010
1:48 AM
|
MP - "5F6H since you got all argy bargy and said "RUBBISH", you're not allowed to join in the single malt tests w/ groyster and Bobo in scotland. so there."
Ha, then I'll keep my 20yr old, single cask, Brora & Caol Isla to myself! :-)
|
MP
731 posts
Jul 19, 2010
1:05 PM
|
now i've done it!
[MP adjusts his pince nez and thinks of a way to apologize to 5F6H without actually apologizing....hmmm?]
----------
MP
hibachi cook for the yakuza
doctor of semiotics
superhero emeritus
|
Brendan Power
56 posts
Jul 20, 2010
3:57 AM
|
A webpage for the SPAH 2010 comb test is now up. Go to
http://www.brendan-power.com/
and click the link in upper left (above the music player).
There are photos of the quick-change bracket Vern Smith has made for masking the comb weights, plus description of procedure etc. The page will be updated periodically before SPAH.
|
Post a Message
|
